Our work on this project includes automated assays for inhibitor screening, cloning and expression of target enzymes, and immunolocalization protocols for both proteases and endogenous inhibitors. We recognize that there may be variation in the type of protease expressed by different tumors, or multiple proteases may be produced by the same tumor. To gain a perspective on protease variation within individual tumors, and to ensure all key targets are examined, we will screen each of our human epithelial cancer tissue banks for proteolytic activity. Automated protease assays will allow detection of individual classes of enzymes in each tumor sample. This screen should provide information on whether certain tumor types will require more than one class-specific inhibitor and whether we should consider targeting other proteases (e.g. cathepsin D, cathepsin L, or other metalloproteases). The automated assays we will use for this retrospective screening of tumor tissue banks may later serve as diagnostic assays to direct inclusion of class-specific inhibitors as adjuvant therapy in individual patients. We will also screen tumor cell lines and confirm protease or inhibitor expression. By comparison to the relevant human cancer tissue bank, the cell lines which best reflect protease expression in actual tumor samples can be identified. Proteases (and, if relevant, protein inhibitors) will be localized by immunohistochemistry or active site-directed biotinylated probes. If any variants of known proteases are identified as having altered substrate specificity or localization, they will be further analyzed by PCR amplification and sequencing. Finally, when tumor samples and cell lines are assayed, aliquots will be provided to Program 3 for assays of multidrug transporter, P-glycoprotein, cytochrome P450s, glutathione-S-transferases and epoxide hydrolases.